-
Notifications
You must be signed in to change notification settings - Fork 0
/
ASL.py
244 lines (213 loc) · 9.22 KB
/
ASL.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
#
# ======================================================
# import libs
# ======================================================
#
# import the lib
from tkinter import *
from PIL import ImageTk,Image
import tkinter.font as tkFont
from tkinter import filedialog
# import nesscary libary
# openCV lib
import cv2 as cv
# matplotlib lib
# numpy lib
import numpy as np
# import os
import os
import tensorflow as tf
from tensorflow import keras
from tkinter import messagebox
from tensorflow.keras import datasets, layers, models
from sklearn.model_selection import train_test_split
import shutil
import webbrowser as wb
root = Tk()
# ini root windows 700x550 pixels
root.title("ASL")
list_nums = []
obj2 = {}
if os.path.exists("./datasets"):
shutil.rmtree("./datasets")
def submit_train_data(data):
global list_nums
list_nums.append(data)
create_dir()
obj2[data] = 0
video = cv.VideoCapture(0)
while(True):
isTrue, frame = video.read()
# flip the image
frame = cv.flip(frame, 1)
# draw rect at top right corner of screen
cv.rectangle(frame, (0,0), (frame.shape[1]//2 - 50, frame.shape[0]//2 - 50), (0,255,0), thickness = 1)
# Write text
cv.putText(frame, 'Mode:collecting images', (0,230), cv.FONT_HERSHEY_SIMPLEX , 1.0, (0,255,0), 2)
cv.putText(frame, 'Count {} : {}'.format(data,obj2[data]), (0,260), cv.FONT_HERSHEY_SIMPLEX , 0.8, (255,0,0), 2)
# show the image
cv.imshow('frame',frame)
k = cv.waitKey(33)
if k == 27: # Esc key to stop
break
elif k == -1: # normally -1 returned,so don't print it
continue
elif k == 32:
print(obj2)
cropped_img = frame[1:(frame.shape[0]//2) - 50,1:(frame.shape[1]//2) - 50]
gray = cv.cvtColor(cropped_img, cv.COLOR_BGR2GRAY)
blur = cv.medianBlur(gray, 7)
th3 = cv.adaptiveThreshold(blur,255,cv.ADAPTIVE_THRESH_GAUSSIAN_C,\
cv.THRESH_BINARY_INV,11,2)
# resizing the image
resized = cv.resize(th3, (100,100), interpolation=cv.INTER_AREA)
# save the crop images to target folder
cv.imwrite('./datasets/{}/{}.jpg'.format(data,str(obj2[data])), resized)
# increase the counter
obj2[data] += 1
if(obj2[data] == 100):
messagebox.showinfo("collecting images", "You have reach 100 images data, now train the model")
video.release()
cv.destroyAllWindows()
def create_dataset(img_folder):
img_data_array=[]
class_name=[]
for dir1 in os.listdir(img_folder):
for file in os.listdir(os.path.join(img_folder, dir1)):
image_path= os.path.join(img_folder, dir1, file)
image= cv.imread(image_path,0)
image=np.array(image)
image = image.astype('float32')
image /= 255
img_data_array.append(image)
class_name.append(dir1)
return img_data_array, class_name
obj3 = {}
def train_model_func():
if not os.path.exists("./datasets"):
messagebox.showerror("Please create datasets", 'Please register new sign to train model')
global list_nums
img_data, class_name =create_dataset('./datasets')
img_data = np.array(img_data)
class_name = np.array(class_name)
global obj3
for i in list_nums:
obj3[i] = list_nums.index(i)
class_name = np.array([obj3[x] for x in class_name])
img_data = img_data.reshape((len(list_nums)*100, 100, 100,1))
X_train, X_test, y_train, y_test = train_test_split(img_data, class_name, test_size=0.3, random_state=101)
model = models.Sequential([
layers.Conv2D(64, 3, padding='same', activation='relu', input_shape = (100,100,1)),
layers.MaxPooling2D(),
layers.Dropout(0.3),
layers.Conv2D(32, 3, padding='same', activation='relu'),
layers.MaxPooling2D(),
layers.Dropout(0.3),
layers.Conv2D(16, 3, padding='same', activation='relu'),
layers.MaxPooling2D(),
layers.Dropout(0.3),
layers.Flatten(),
layers.Dense(128, activation='relu'),
layers.Dense(64, activation='relu'),
layers.Dense(len(list_nums))
])
model.compile(optimizer='adam',
loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
metrics=['accuracy'])
history = model.fit(X_train, y_train, epochs=10, validation_data=(X_test,y_test))
# win2 = Toplevel()
# win2.title('Second windows')
# bb = Button(win2, text="Create new window")
# bb.pack()
acc = model.evaluate(X_test,y_test)
model_filename = filedialog.asksaveasfilename(initialdir="./", title="Save A File", filetypes=(("h5 files", "*.h5"),("all files", "*.*")))
model.save(model_filename)
def new_windows():
top = Toplevel()
top.title('Add new gesture')
fontStyle2 = tkFont.Font(family="Lucida Sign", size=15)
title_ASL2 = Label(top, text = 'Add new Sign', font = fontStyle2)
new_entrry = Entry(top, text = "Please enter new sign")
button_sign = Button(top, text = "Submit", command = lambda: submit_train_data(new_entrry.get()))
title_ASL2.grid(row = 0, column = 0, columnspan = 1)
new_entrry.grid(row = 1, column = 0)
button_sign.grid(row = 1, column = 1)
def create_dir():
global list_nums
if not os.path.exists("./datasets"):
os.makedirs("./datasets")
for i in list_nums:
if not os.path.exists(f"./datasets/{i}"):
os.makedirs(f"./datasets/{i}")
def process_img(frame):
cropped_img = frame[1:(frame.shape[0]//2) - 50,1:(frame.shape[1]//2) - 50]
gray = cv.cvtColor(cropped_img, cv.COLOR_BGR2GRAY)
blur = cv.medianBlur(gray, 7)
th3 = cv.adaptiveThreshold(blur,255,cv.ADAPTIVE_THRESH_GAUSSIAN_C,\
cv.THRESH_BINARY_INV,11,2)
# resizing the image
resized = cv.resize(th3, (100,100), interpolation=cv.INTER_AREA)
return resized
obj = {0:'0',1:'1',2:'2',3:'3',4:'4',5:'5', 6:'6', 7:'7', 8:'8', 9:'9',10:'A',11:'B',12:'C',13:'D',14:'E',15:'F',16:'G',17:'H',18:'I',19:'J',20:'K',21:'L',22:'M',23:'N',24:'O',25:'P',26:'Q',27:'R',28:'S',29:'T',30:'U',31:'V',32:'W',33:'X',34:'Y',35:'Z',36:' '}
def prediction_clicked():
global obj
video = cv.VideoCapture(0)
root.filename = filedialog.askopenfilename(initialdir="./", title="Select A File", filetypes=(("h5 files", "*.h5"),("all files", "*.*")))
last_path = os.path.basename(os.path.normpath(root.filename))
if(last_path != 'model_ASL.h5'):
obj = dict(zip(obj3.values(), obj3.keys()))
loaded_model = keras.models.load_model(root.filename)
while(True):
# read the video images
isTrue, frame = video.read()
# flip the image
frame = cv.flip(frame, 1)
# draw rect at top right corner of screen
cv.rectangle(frame, (0,0), (frame.shape[1]//2 - 50, frame.shape[0]//2 - 50), (0,255,0), thickness = 1)
# Write text
cv.putText(frame, 'Mode:predict model', (0,230), cv.FONT_HERSHEY_SIMPLEX , 1.0, (0,255,0), 2)
# process img
process_images = process_img(frame)
# predit image
text = np.argmax(loaded_model.predict(process_images.reshape(-1,100,100,1)), axis=-1)
cv.putText(frame, 'Prediction: {}'.format(obj[text[0]]), (0,290), cv.FONT_HERSHEY_SIMPLEX , 1.0, (255,0,0), 2)
cv.imshow('frame',frame)
cv.imshow('Threshold image',process_img(frame))
k = cv.waitKey(33)
if k == 27: # Esc key to stop
break
video.release()
cv.destroyAllWindows()
#
# ======================================================
# init widgets
# ======================================================
#
fontStyle = tkFont.Font(family="Lucida Grande", size=30)
fontStyle2 = tkFont.Font(family="Lucida Grande", size=12)
title_ASL = Label(root, text = 'Sign lanaguge trasnaltor', font = fontStyle)
subtitle = Label(root, text = 'SLT is tool that uses computer vison to track a users gestures', font = fontStyle2)
subtitle2 = Label(root, text = 'And then used a learned model to identify the SLT character most ', font = fontStyle2)
subtitle3 = Label(root, text = 'correlated to that gesture.', font = fontStyle2)
register_sign = Button(root, text = 'Register new sign', padx = 10, pady = 10, command = new_windows)
train_model = Button(root, text = 'Train new model',padx = 10, pady = 10, command = train_model_func)
make_prediction = Button(root, text = 'Make prediction',padx = 10, pady = 10, command = prediction_clicked )
my_img = ImageTk.PhotoImage(Image.open("./images/a.jpg"))
maj_logo = Label(root, image = my_img)
ai_image = ImageTk.PhotoImage(Image.open("./images/artifical_inteligence2.png"))
ai_logo = Label(root, image = ai_image)
#
# ======================================================
# Grid System
# ======================================================
#
maj_logo.grid(row = 0, column = 0, columnspan = 3)
title_ASL.grid(row = 1, column = 0, columnspan = 3 )
subtitle.grid(row = 2, column = 0, columnspan = 3 )
subtitle2.grid(row = 3, column = 0, columnspan = 3 )
subtitle3.grid(row = 4, column = 0, columnspan = 3 )
ai_logo.grid(row = 5, column = 0, columnspan = 3 )
register_sign.grid(row = 6, column = 0)
train_model.grid(row = 6, column = 1)
make_prediction.grid(row = 6, column = 2)
root.mainloop()